This project addresses the automatic control of general anesthesia during surgery, with the objective of developing an efficient and robust solution, to increase safety and reduce post-operative complications. The project aims to design a controller that determines the rate of the hypnotic and analgesic agents to achieve a rapid anesthesia onset and maintain the BIS signal in a safe range. The control scheme will be based on robust model predictive control.
Sito Web: https://activa.unibs.it/
1) Improve safety: the patient's level of sedation is constantly monitored during surgery and intensive care.
2) Avoid drug over/underdose: the BIS is always maintained in a safe range.
3) Reduce wake-up time: the patient awakes quickly and without side effects from excessive sedation.
4) Reduce workload for the anesthesiologist: he/she does not need to constantly monitor the patient’s BIS and has more time for other tasks during surgery.
The control will minimize the time required for inducing unconsciousness, avoiding excessive undershoot of the BIS signal. The controller is designed to maintain the BIS signal in a prescribed range, despite the presence of noxious stimuli and the uncertainty in model parameters. The devised methodologies will be tailored to critically ill patients in ICUs, for which control requirements are tighter.
We identified pharmacokinetic and pharmacodynamic models of patients from data available during surgery. We optimized the co-administration of propofol and remifentanil during surgery.
We used both PID control and model predictive control for drug coadministration on identified patient models. We optimized the set-point concentration of propofol and remifentanil in target controller infusion. At Brescia hospital, we performed clinical trials on patients during surgery.
